论文信息 - REAL-TIME ONSET DETECTION WITH RECURRENT NEURAL NETWORKS

REAL-TIME ONSET DETECTION WITH RECURRENT NEURAL NETWORKS

We present a new onset detection algorithm which operates online in real time without delay. Our method incorporates a recurrent neural network to model the sequence of onsets based solely on causal audio signal information. Comparative performance against existing state-of-the-art online and offline algorithms was evaluated using a very large database. The new method – despite being an online algorithm – shows performance only slightly short of the best existing offline methods while outperforming standard approaches.

Florian Krebs | Markus Schedl | Sebastian Böck | Andreas Arzt

[1] S. Dixon. ONSET DETECTION REVISITED , 2006 .

[2] Nick Collins. A Comparison of Sound Onset Detection Algorithms with Emphasis on Psychoacoustically Motivated Detection Functions , 2005 .

[3] Björn W. Schuller,et al. Universal Onset Detection with Bidirectional Long Short-Term Memory Neural Networks , 2010, ISMIR.

[4] Yannis Stylianou,et al. Three Dimensions of Pitched Instrument Onset Detection , 2010, IEEE Transactions on Audio, Speech, and Language Processing.

[5] Mark B. Sandler,et al. A tutorial on onset detection in music signals , 2005, IEEE Transactions on Speech and Audio Processing.

[6] Douglas Eck,et al. A Supervised Classification Algorithm for Note Onset Detection , 2006, EURASIP J. Adv. Signal Process..

[7] Dan Stowell,et al. Adaptive whitening for Improved Real-Time audio onset Detection , 2007, ICMC.

[8] Matthew E. P. Davies,et al. Onset Event Decoding Exploiting the Rhythmic Structure of Polyphonic Music , 2011, IEEE Journal of Selected Topics in Signal Processing.